Those old enough to have encountered punch cards in their lifetime are probably glad to be rid of their extremely low data density and the propensity of tall stacks to tip over. But obsolete as they may be, they’re a great tool to show the basics of binary data storage: the bits are easily visible and can even be manipulated with simple tools. As an experiment to re-create those features in a more modern system, [Michael Kohn] made a punch card-like system based on LEGO bricks that stores machine code instructions for a 65C816 CPU, the 16-bit successor to the venerable 6502.
Bits are stored on a white 8×20 stud board, onto which small black pieces are placed. A white background stud encodes a logic “zero”, while a black stud encodes a logic “one”. The bits are read out by an array of reflectance sensors, which conveniently has the same 8 mm pitch as standard LEGO studs. A big wheel driven by a stepper motor slides the data card under the readout circuit along a short stretch of LEGO train tracks.
The optical sensors are read out by an MSP430 series microcontroller, which also drives the motor through a stepper motor driver. Once the data is read out, the bytes are transferred into a WDC W65C265SXB board, which executes them as machine code instructions on its 65C816 CPU. In the video below, you can see a program being loaded that blinks an LED.
As part of [Erich Styger]’s recent kitchen overhaul with more power-efficient machines, he came across the ‘AutoDos’ feature of the new Miele G 27695-60 dishwasher. These are essentially overpriced containers of dishwashing powder that go into a special compartment of the machine, from which the dishwasher can then dispense the powder as needed. The high price tag and purported single-use of these containers led to the obvious question of whether they can be refilled.
With a cost of over $10 per PowerDisk container, each containing 400 grams of powder that suffice for ~20 cycles, it should be obvious that this is not a cheap system. Fortunately, each PowerDisk is just a foil-covered plastic container with no real special components. This meant that one hole and some funnel action later, [Erich] had refilled an empty PowerDisk with fresh powder, with the Miele dishwasher happily purring away and none the wiser that it wasn’t using Genuine Miele PowerDisk Dishwasher Powder™.
How well this system holds up long-term is uncertain, as the containers were not designed for constant reuse, but it offers the perspective of some creative 3D printing to create an (ABS-based?) container alternative. Having a automatic powder or liquid dosing system in a dishwasher is a pretty useful feature, but when it gets tied to what is clearly a cash grab, it rather ruins the deal.
“From scratch” is a bit of a murky expression. How scratchy does it get? Are you just baking your bread yourself or are you growing your own wheat? Rolling your own solar installation probably doesn’t involve manufacturing your own photovoltaic cells. But when it comes to making the frame to hold your panels, why not machine your own brackets and harvest the wood from trees nearby?. That’s what [Kris Harbour] did with his over-engineered 8.4kW solar stand.
He already had a wind turbine and a few solar panels elsewhere on the property, but [Kris] wanted a bit more power. At the start of the video, [Kris] makes an offhand comment that he wants this to last 30-40 years at a minimum. Everything from the focus on the concrete footings to the oversized brackets and beams reflects that. The brackets were cut on his plasma CNC and the wood was cut to rough dimensions using a sawmill on his property. He admits that the sapwood will rot away a bit after 20ish years, but since the core of the posts is heartwood, structurally they’ll last a long time. After planing down to the right size, cutting hole and grooves was a bit of an involved procedure. The structure is quite tall once erected and we loved watching it come together. The most impressive thing here is that this structure was put together by just a single person. All the rafters were cut and hand-chiseled to the right sizes and hoisted into place. The panels ultimately weren’t delivered on time and he had to switch to using new panels rather than the used panels he was expecting.
Previously, [Kris] had shown off his impressive hydroelectric setup. We’re looking forward to seeing the new solar array power all the projects [Kris] is planning in the future. Video after the break.
The inconvenience of having to walk to your mailbox to check for mail has inspired many hackers to install automated systems that let them know when the mail has been delivered. Mailbox monitors have been made based on several different mechanisms: some measure the weight of the items inside, some use cameras and machine vision, while others simply trigger whenever the mailbox’s door or flap is moved. When [Gary Watts] wanted to install a notification system for his 1940s brick letterbox, his options were limited: with no flap or door to monitor, and limited space to install mechanical contraptions, he decided to use a LIDAR sensor instead.
Probably best-known for their emerging application in self-driving cars, LIDAR systems send out a laser pulse and measure the time it takes for it to be reflected off a surface. In the case of [Gary]’s mailbox, that surface is either the brick wall or a letter leaning against it. Since letters are inserted through a vertical slot, they will usually be leaning upright against the wall, providing a clear target for the laser.
The LIDAR module, a VL53L0X made by ST, is hooked up to a Wemos D1 Mini Pro. The D1 communicates with [Gary]’s home WiFi through an external antenna, and is powered by an 18650 lithium battery charged through a solar panel. The whole system is housed inside a waterproof plastic case, with the LIDAR sensor attached to the inside of the mailbox through a 3D-printed mounting bracket. On the software side, the mailbox notifier is powered by Home Assistant and MQTT. The D1 spends most of its time in deep-sleep mode, only waking up every 25 seconds to read out the sensor and send a notification if needed.
There has been a rumor that Apple is working on a glucose monitoring solution for the Apple watch. [Harley] decided not to wait and managed to interface an Abbot FreeStyle Libre sensor with the Apple watch. The sensor doesn’t directly read glucose continuously, but it does allow for more frequent reading which can help diabetic patients manage their blood sugar levels. However, as part of the hack, [Harley] effectively converts the meter to a continuous-reading device, another bonus.
The trick is to add a Bluetooth transmitter to the NFC sensor. Using a device called a MiaoMiao, the task seems pretty simple. The MiaoMaio is small, waterproof, and lasts two weeks on a charge, which is longer than the sensor’s life. Honestly, this is the hack since once you have the data flowing over Bluetooth, you can process it in any number of ways including using an app on the Apple watch.
It isn’t perfect. There’s a slight lag with readings due to the way the sensor works. However, you usually don’t care as much about the absolute value of your glucose (unless it is very high or very low). You are usually more interested in the slope of the change. This data is more than good enough for that.
In fact, the most complex part of this seems to be the watch app. It might be less work to feed the data to a machine learning model and let AI guide your insulin injections. Something to think about.
We have a keen interest in glucose monitoring around here and we know why it is so darn hard. Honestly, the idea of pushing glucose meter data to a watch isn’t new, but this is a well-done implementation with a lot of possibilities.
The community of Hackaday readers is diverse and talented, and supplies us with plenty of motivation, feedback, knowledge, and of course cool stuff to show you. There are many interest streams within it, but it’s safe to say that we’re more directed towards the hardware scene here. One of those parallel streams which has much overlap is the demoscene, that area in which programming, art, and music come together and push computer hardware to the limit of its abilities in pursuit of the most eye-catching works. I took a road trip with a friend to Outline, a small demo party held on a farm in the eastern Netherlands, to take a look at the world of demos up-close as a hardware-focused outsider.
Like A Hacker Camp, But The Music’s Better
A still from Thrive, a 256 byte demo for the TIC-80 fantasy console.
If I wanted to sum up the flavour of Outline, I’d describe it as very similar to a small hacker camp, but with better music and partying. The hackerspaces are replaced by demo groups and awesome graphics take the place of robots and electronics, but the vibe of people with a passionate interest in the low-level understanding of technology is exactly the same. Even some of the same faces make an appearance. On the benches sit modern high-spec PCs alongside classic consoles and microcomputers, on the projector screen are live coding shaders or some of the most recognisable demos past and present, and in the air is an eclectic mix of live-DJ EDM and chiptunes.
As an outsider at a first demoscene event it’s difficult to appreciate the work from a comparative perspective, while like most of us I’m familiar with quite a few demos that have become popular I’m not well-equipped enough to talk about the code and techniques behind them But I can run through the various sections of the competition, and since everything is online I can link to a few of them. The competition is split up into several sections, which are loosely for all-out technology-no-object demos, space-limited 256 byte and 128 byte demos, and old-school demos for retrocomputing hardware. Each is a test of the programmer’s skill in fitting the most into the least of resources, and for those who appreciate such things it’s the cleverness of the technique which produces the demo that’s as much a draw as the look of the thing. I don’t think I have ever exercised such mastery over any of the computers I have owned. So browse the entries, and marvel at their ingenuity. My personal aesthetic favourites were Thrive by [Agenda] for the TIC-80 fantasy console and It’s about time by [Guideline] for Windows, but you may have different tastes.
Don’t Forget The Hardware
Mine Storm 4D, on a lenticular holographic display.
Beyond the atmosphere and the demos themselves, there was a bit of hardware for the retrocomputer enthusiast. The Atari Falcon and Jaguar were neither destined to set the world on fire when they appeared, but there they were for those of us who drooled over them back in the day to lust for once more.
If the original hardware wasn’t enough then there was some newly minted retrocomputing hardware making a showing, with a couple of minimig Amiga FPGA boards showing Workbench. Star of the hardware show though went to Mine Storm 4D, a version of the classic Vectrex game Mine Storm running on a PC, for the Looking Glass Factory holographic portrait display. With my visual superpower I didn’t quite get a 3D effect, but I definitely got the holographic effect when moving my head.
Having never been to a demo party I didn’t know quite what to expect, but I can safely say I had a fantastic time, saw a lot of really cool stuff, and made some friends along the way. If you’ve never been to a demo party because it’s not quite your scene then all I can say is that you should give it a go. Every hardware hacker should go to a demo party!
When it comes to robots, especially ones that need to achieve some degree of autonomy, the more constrained the environment they work in, the easier it is for them to deal with the world. An industrial arm tethered next to a production line, for example, only has to worry about positioning its tool within its work envelope. The problems mount up for something like an autonomous car, though, which needs to deal with the world in two — or perhaps two and a half — dimensions.
But what about adding a third dimension? That’s the realm that aerial robots have to live and work in, and it’s where the problems get really interesting. Not only are there hardly any constraints to movement, but you’ve also got to deal with the problems of aerodynamic forces, navigation in space, and control systems that need to respond to the slightest of perturbations without overcompensating.
The atmosphere is a tough place to make a living, and dealing with the problems of aerial robotics has kept Nick Rehm occupied for many years as a hobbyist, and more recently as an aerospace engineer at Johns Hopkins Applied Physics Laboratory. Nick has spent his time away from the office solving the problems of autonomous flight, including detection and avoidance of mid-air collisions, development of vertical take-off and landing (VTOL) and fixed-wing aircraft, and even ground-effect aircraft. He’ll drop by the Hack Chat to discuss the problems of aerial robots and the challenges of unconventional aviation, and help us figure out how to deal with the third dimension.
